[Bug runtime/17126] tracepoints.exp testcase causing stalls/hang on ppc64

["dsmith at redhat dot com" <sourceware-bugzilla at sourceware dot org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=17126

--- Comment #1 from David Smith <dsmith at redhat dot com> ---
>From some more debugging, it appears to be something going on with tracepoints
that start with 'h'. If I change tracepoints.stp to probe
'kernel.trace("[a-gi-z]*")', the testcase works fine.

Here are the tracepoints that start with 'h':

====
# stap -l 'kernel.trace("h*")'
kernel.trace("hcall_entry")
kernel.trace("hcall_exit")
kernel.trace("host1x_cdma_begin")
kernel.trace("host1x_cdma_end")
kernel.trace("host1x_cdma_push")
kernel.trace("host1x_cdma_push_gather")
kernel.trace("host1x_channel_open")
kernel.trace("host1x_channel_release")
kernel.trace("host1x_channel_submit")
kernel.trace("host1x_channel_submit_complete")
kernel.trace("host1x_channel_submitted")
kernel.trace("host1x_syncpt_load_min")
kernel.trace("host1x_syncpt_wait_check")
kernel.trace("host1x_wait_cdma")
kernel.trace("hrtimer_cancel")
kernel.trace("hrtimer_expire_entry")
kernel.trace("hrtimer_expire_exit")
kernel.trace("hrtimer_init")
kernel.trace("hrtimer_start")
====

So, there are basically 3 sets: hcall_*, host1x_*, and hrtimer_*. Each of those
sets works fine individually. I haven't quite found the combination that causes
the stall.

-- 
You are receiving this mail because:
You are the assignee for the bug.